Conscripting Terahertz Sensors

Lynn Savage, Features Editor, lynn.savage@photonics.com

Imaging and spectroscopy in the terahertz frequency range eventually will give troops a welcome advantage in the field.

Terahertz waves are short enough to provide resolution of less than 1 mm, yet long enough to penetrate most nonmetallic substances, such as the materials used to make clothing, rucksacks and tarps. As such, they are useful to security agents and military personnel alike for revealing concealed weapons, chemical explosives and biological agents. Besides security applications such as airport screeners, higher-resolution terahertz sensors could provide enhanced identification of battlefield targets, better missile guidance and other combat advantages.

Soldiers, marines and fighter pilots are increasingly trained to use not only the visible wavelengths that their eyes can process, but infrared wavelengths as well. To some in the military-aligned industries, sensors that read terahertz frequencies could help augment what the US Army calls its Future Combat System by supporting a wide range of microdevices that scan multiple frequencies at all times.

The US Army Research Office is a main driver of all terahertz-related defense technologies, according to Dwight Woolard of the US Army Research Laboratory in Research Triangle Park, N.C. However, he noted, because the terahertz spectrum is very broad with extreme diversity across the regime, the ARO places emphasis on select scientific and technology projects that are high risk and high reward.

As far as researchers in the field are concerned, the terahertz range runs from 300 GHz to as high as 10 THz. The spectral territory beyond that is largely unexplored until you get closer to microwave frequencies.

Terahertz technology looks at the “collective motion of molecules stacked in a group,” said M. Hassan Arbab of the University of Washington in Seattle. “The spectral lines come from the vibrational modes of large molecules.”

Spectral information acquisition has thus far relied upon only a couple of basic technologies, namely antenna-coupled semiconductors, cryogenically cooled bolometers and uncooled microbolometers paired with quantum-cascade lasers (QCLs).

Antenna-coupled sensors operate in the subterahertz (0.1 to 1 THz) or low-terahertz range (<2 THz), whereas thermal detectors provide broader spectral coverage, depending upon the characteristics of the terahertz absorber employed, said Gamani Karunasiri of the Naval Postgraduate School (NPS) in Monterey, Calif.

Terahertz imaging reveals a surgical blade hidden inside a piece of Styrofoam. Courtesy of Gamani Karunasiri, Naval Postgraduate School.

“Due to the lack of appreciable terahertz power in the thermal background, it is necessary to use an illumination source for imaging using uncooled thermal sensors,” Karunasiri said. “Initial terahertz imaging using a standard infrared microbolometer camera with modified optics and a quantum cascade laser illuminator showed good promise in real-time imaging.”

Rough work

Making an explosive may not actually resemble cooking in the kitchen, but conceptually you end up with something akin to a meatloaf. As with the chunks of hamburger and breading that make the meal, explosives such as C-4 and RDX (cyclotrimethylenetrinitramine, or cyclonite) are composed of a mash of volatile chemicals and binders.

Capturing reflectance from such mixtures requires a certain amount of surface roughness in the target material. Plastic explosives such as C-4, for example, have fairly rough particles, which have the same approximate size as the terahertz waves themselves. However, particles over a certain size tend to overly scatter terahertz waves, affecting overall absorption. Therefore, despite being mostly acceptable targets of terahertz radiation, some targets will have particles that naturally dampen the ability of a scanner to successfully identify them.

Measured absorption characteristics of a metamaterial structure and THz-QCL emission spectrum. Courtesy of Gamani Karunasiri, Naval Postgraduate School.

Oddly, alpha-lactose, a form of the common sugar found in milk, makes a handy (and safe) stand-in for some explosives. Particles made of the materials, when excited, exhibit resonances at 540 GHz, 1.2 THz and 1.38 THz. Pellets made chiefly of lactose particles are available in a multitude of sizes representing various surface roughness grades (akin to sand-paper ratings), and tests using these have shown that terahertz waves become less reliable as the particle size – and thus overall surface roughness – increases.

In 2010, researchers at the University of Washington devised a mathematical transform method to address the issue of surface roughness. Typical time-domain spectrographic analysis involves acquiring multiple pulsed reflection measurements from several spots on the target surface, then calculating the Fourier transform of each metric. At terahertz frequencies, however, target irregularities scatter so much energy that any signal gets buried in an avalanche of noise.

Lactose pellets stand in for more volatile chemicals when testing the effects of surface roughness on terahertz sensing. a 5 400 grit (23-µm particle diameter); b 5 150 grit (92 µm); c 5 P80 grit (201 µm). (d) shows the normalized reflection spectral amplitudes of these samples. Reprinted with permission of Applied Physics Letters.

Instead, the Washington group – Arbab and his colleagues Antao Chen, Eric I. Thorsos and Dale P. Winebrenner – tweaked a wavelet transform method, not uncommon in terahertz investigations. Dubbed the maximal overlap discrete wavelet transform, the group’s technique permits a better spectral signal-to-noise ratio from targets, even given very rough surface equivalents and only a few disjoint terahertz measurements of the target particles. The researchers report, however, that more work is needed to nail down the minimum number of measurements needed for certainty.

Their development of the wavelet transform was reported in Proceedings of SPIE, Vol. 7601, and in Applied Physics Letters, Vol. 97, 181903.

Illumination

“Since terahertz radiation is highly absorbed by the atmosphere, long-range standoff imaging requires relatively high-power terahertz illuminators, depending on the distance,” said Karunasiri of the NPS. “The best laser sources available to date are either terahertz QCLs or free-electron lasers (FELs), which require [a] large infrastructure to operate.”

FELs offer high power and a wide range of tunability within the terahertz frequency range, which is important not only for imaging, but also for spectroscopy, which permits better identification of chemical and biological threats. More compact FELs that are under development should lead to compact high-power sources for field applications, Karunasiri said.

Graphene – individual sheets of carbon atoms – also is gaining attention as a possible terahertz source. The optoelectronic properties of graphene make the material an interesting subject for a number of research groups, including that of Alexander W. Holleitner at the Technical University of Munich in Garching, Germany.

When pumped by a QCL laser, two or more layers of graphene freely suspended on bimetal supports emit terahertz radiation. This could lead to compact emitters necessary for portable operation. Courtesy of Alexander W. Holleitner, Technical University of Munich.

“Generally, my group explores the time-resolved optoelectronic dynamics in a variety of nanoscale materials, including nanowires, carbon nanotubes, metal nanoparticles, molecules and corresponding hybrid structures,” Holleitner said. “The dynamics comprise ultrafast dielectric displacement currents, the drift currents of photogenerated charge carriers and recombination processes.” Furthermore, the tunable bandgap in bilayer graphene is in the terahertz range, he said.

Holleitner and his colleagues reported in the April 12, 2011, issue of Nature Communications that they tested the photoelectric response of graphene. They placed bilayered graphene into a circuit that also included a pair of titanium and gold striplines.

Using a mode-locked Ti:sapphire laser with a pulse width of ~160 fs and a repetition rate of about 76 MHz, the investigators pumped the graphene. Terahertz radiation was emitted from a charge-carrier plasma that was generated by the optically pumped graphene. Increasing the laser power, Holleitner said, increased the frequency of the terahertz emission.

“The striplines are made out of titanium and gold, but in principle, any metal will do it,” Holleitner said. The co-planar striplines fulfilled two purposes for the researchers. First, they acted as a near-field antenna for the terahertz radiation; and second, they transduced the signal to an ultrafast low-temperature-grown GaAs photodetector placed to the side of one of the metal striplines. The stripline circuit is up to five orders of magnitude more sensitive in picking up the terahertz signal than are far-field detection mechanisms, Holleitner said.

Attenuation due to atmospheric effects is a problem. Attenuation can be hundreds of decibels per kilometer. At shorter distances – say, hundreds of meters – it is less of an issue. Large bands in the 300-GHz to 2-THz and 6- to 10-THz ranges are particularly less prone to atmospheric effects. More powerful lasers, such as free-electron devices, can increase the amount of safe distance, but are so large as to preclude field use.

Diode lasers are more compact, of course, but are generally very low in power. This is a trade-off that researchers are attempting to avoid with new technologies.

For example, one alternative to addressing the attenuation problem has been proposed: Xi-Cheng Zhang of the University of Rochester in New York and his colleagues have developed a method to generate terahertz waves close to a target, then analyze and communicate the spectroscopic findings to personnel a safe distance away via 800-nm waves.

The future warrior

The most important terahertz application, Karunasiri said, is standoff detection of concealed weapons and identification of explosives. At the NPS, work on future terahertz sensors includes the development of thermal detectors based on microbolometers and on “bimaterial pixels.”

Bimaterial detectors, which combine heat-sensitive substances such as SiO₂ and aluminum, rely on the deflection resulting from differences in the thermal expansion of the two materials, Karunasiri said. Besides focal plane arrays comprising SiO₂ and aluminum (sans substrate), his group has developed nanoscale films made of chromium or nickel that provide up to 50 percent terahertz absorption within the 1- to 10-THz range. The film thickness and conductivity play an important role in achieving the 50% absorption for a given metal. The group has also developed metamaterial-based narrowband (~1 THz) devices that absorb nearly 100 percent at a targeted terahertz frequency in the 1- to 10-THz range.

A schematic (below) and micrograph (above) show the design and fabrication of a metamaterial structure tuned to a quantum cascade laser (QCL) frequency. Courtesy of Gamani Karunasiri, Naval Postgraduate School.

“Our goal is to develop highly sensitive terahertz focal plane arrays for real-time terahertz imaging,” he said.

“Everything is 10 to 20 years down the road,” Arbab said. But one day, robots will roll down the road, emitting laser beams toward suspected targets and identifying dangerous contents within innocuous-looking packages, he suggests. Currently, terahertz technology can be used to look for the spectral signatures of C-4, TNT, RDX and more. A spectral database of most/all of the known explosives is “practically there,” he said.

It’s now a matter of making the technology catch up.

The Main Hurdles to Terahertz Technology

#1. Terahertz imaging and terahertz spectroscopy use different partsof the spectrum and require different technologies. Imaging at 300 GHz (as with airport scanners) is not harmful to people, but you don’t get vital spectrographic information from it.

#2. Water vapor, even in very dry deserts, brings about major atmospheric attenuation; for example, in a room with 20 percent humidity, a terahertz signal ends at 1 m.

#3. More powerful lasers are needed to make both terahertz spectroscopy and imaging possible from greater distances between source and target, and powerful QCLs and FELs are too large and expensive to hand out to troops in the field.

Applying Terahertz Waves to Burn Triage

Terahertz radiation, it turns out, isn’t only good for determining the threat level of a would-be terrorist or combatant.

“The problem with military/security applications is that you’re always playing hide-and-seek,” said M. Hassan Arbab of the University of Washington in Seattle. “It’s important, but it occurred to us that there are targets that don’t hide.”

More than a million people are treated for burns in the US every year; battle wounds and burns also must be evaluated to determine which tissue is damaged, how badly it is damaged and which is still healthy. Third-degree burns are the most problematic because they require excision of the injured tissue. It’s not as easy as making a quick visual inspection; current techniques for evaluating third-degree burns are only about 60 percent successful, and that is under optimal conditions.

Terahertz radiation, which is nonionizing and thus will not further harm skin cells, is highly sensitive to the water content of the dermal layers of the skin. When heat burns the skin, fluids build up between the skin cells, protecting what’s left of the dermis. Arbab and his colleagues reported in the Aug. 11, 2011, issue of Biomedical Optics that terahertz signals increase when the fluids – called interstitial edema – are present. They found, for example, that third-degree burns have about 30 percent higher reflectivity in the terahertz range than does normal skin.

The variations in terahertz reflectivity readings permit rapid diagnosis of not only the current state of a burn, but also of its future. “Terahertz radiation helps tell which partial-thickness burns will progress to full-thickness,” Arbab said.